Whistler began this plate, one of his most complex in terms of state changes, with the majority of work in pure etching. He then used drypoint to define and redefine details within the doorway, including the poses and costumes of the girl on the step in front of it and the woman behind her in the shadows.
Wedmore commented dismissively, 'There are one or two undesirable trial proofs, with much less work in the small spaces between the arches high on the house front, and with the whole effect thinner and weaker.' 11
11: Wedmore 1886 A[more] (cat. no. 154).
As the plate progressed, Whistler added drypoint shading to the water and experimented with roulette, apparently in an attempt to create a broad area of intermediate tone. This apparently led to some misunderstanding by viewers in 1880. The critic of the British Architect commented:
'"The Doorway" has a deal of interesting work in it ... The upper part of the plate, the buildings, figure, &c., are all etched in the ordinary manner, but the water below the doorway ... seems to have been produced with a mezzo-tint tool, that is, as the engravers say, a "ground" has been laid out of which the lights have been either scraped or wiped.' 12
The roulette eventually wore down and was obsured by plate tone, as he turned to inking and wiping to produce the effect of the water in the canal. Drypoint was also added more heavily in the final states, to indicate reflections and to create shadows.
In another apparent experiment in regularizing tonal areas in the plate, Whistler painted or spread acid over the grillework at the top and right of the doorway, producing greyish shading over those parts of the plate. He then burnished highlights along and within the grillework. Whistler was aware of the tonal possibilities of 'open bite', the application of acid directly to the copper plate, from the time of Street at Saverne, when he used it to produce regular tone behind the lines of the sky. However, the effects created by direct application of acid on the plate in The Doorway represent his most experimental use of that method. It was also the last time he sought to regularize broad areas of tone in that way - or, for that matter, with roulette work. In the intaglio work that followed, he achieved a range of tonal effects by manipulating ink on plate surfaces and through dense and intricate networks of etched and drypoint lines.
The Doorway had a stormy history. It went through many states - twenty in all - and Whistler pulled a few proofs for himself from each state. Substantial changes were made to the figures and to the interior.
Whistler himself kept trial proofs of many of the states of this etching. The first of these - reproduced below - is touched with ink and wash ().
Both Whistler and the Fine Art Society kept a record of impressions as they were delivered for the 'First Venice Set'. Whistler noted ten impressions on 16 February 1881, eight on 6 April, three on 25 August and another five on 31 December. 13
The first eight states were printed between 1880 and 1881. They include impressions in black ink on ivory laid paper (); on ivory laid with Arms of Amsterdam watermark (); buff wove (); cream laid paper with Pro Patria watermark (). Of the seventh state, one is in dark brown ink on Japanese paper () and one in black on ivory Japan (), while others are in black ink on western papers, both ivory wove () and laid (). Eighth state impressions are also in dark brown but mostly in black ink (i.e. ).
13: Whistler, 31 December 1881 and 7 January 1882, GUW #01137; E. G. Brown to Whistler, 31 December 1881, #01136.
Whistler printed a proof (and had it 'delivered to Lady Clementina Mitford') on 24 February 1883, and recorded another two on 7 April. On 12 June the F.A.S. asked for one to complete an order for Thomas Way
(1837-1915). Whistler sent another four on 18 June, three on 13 July and two on 16 July 1883. In 1884 he managed two proofs on 6 March. In 1885 he delivered ten on 29 July. These may have been the eleventh state, printed in dark brown ink on both Asian () and western () laid papers. Deliveries were rare thereafter; he sent three on 13 January 1887, twelve on 25 June 1887, and the same again on 2 April 1889. 14
A large group of impressions of the 14th state may date from 1887. They were mostly printed in dark brown ink on laid paper in shades of ivory, cream, off-white and buff, one with a 'WW' countermark () and another with the watermark 'FELLOWS / 1804' () but some were on ivory Japanese (, ).
14: A. H. Dunn to Whistler, 12 June 1883, GUW #01160; [9 August 1883], #12991; F.A.S. account, 20 December 1888, #01217.
There seems to have been a gap in the records or in actually printing The Doorway
in 1890-1891. By June 1892 Whistler was living in Paris. He reckoned - wrongly - that he only had to print 16 more impressions for the F. A. S., and mentioned to Marcus Bourne Huish
(1843-1904): 'Also I have printed a proof extra of the finest set each for the British Museum - ... and of most of the important ones to present to the Librairie National - here - and to King [of] Spain - & Italy -' 15
He then sent five packages of various prints by post to London, but, as he said: 'Of course there still remains a matter of printing from the two plates - "The Doorway" & "the Beggars" ... This when I get by & bye the time and am well in my new studio -' 16
When these etchings were damaged by Customs officers, Whistler - perhaps understandably - used it as an excuse to prevaricate even more:
15: To M. B. Huish, 22 June 1892, GUW #01246, #01245.
16: Whistler to Huish, [29 June 1892], GUW, #02949.
'even were I able to begin again at once, it would take endless trouble & much time to get those two plates again into the condition in which they will give proofs of the same beauty as those harmed … when great labour & much thought has been expended in getting a plate into perfect condition for the press, it is wise that such plate should be printed right out - in spite of the momentary want of popularity.'
Ernest George Brown
(1853/1854-1915) continued to beg Whistler to complete the editions of The Doorway
for the F. A. S. Out of the 100 required, they had received 78 impressions. However, Whistler said he was too busy to print: 'All this great over refinement and anxiety for the perfection of quantity is dead loss & vexation -' he wrote, 'for my own reputation, the complete beauty of one proof is enough - & the production of scores of the same plate is really a madness!'
Brown was placatory, praising Whistler's perfectionism, and was rewarded with another packet of trial proofs, within the month. 19
18: Whistler to Brown, [16/19 May 1893], GUW #03612.
19: Brown to Whistler, 20 May 1893, GUW #01265.
In June of the following year, Huish enquired: 'whether you could be so good as to print us just two copies of the Doorways ... if you cannot manage any more - We have made two sales of these to customers who worry our lives out for them - ' 20 Two month later, Whistler sent 'four superb proofs', including two of The Doorway on 'most "fragile"' Japanese paper, which must be 'pressed again at once - but with such care! -'. He described the impressions as 'Proofs ... final state', but it is not known if these actually were the final state of the etching. 21 They were fine impressions, set aside, the artist said, for '"Crowned Heads" or Museums!' One may have been a superb impression of the final state printed in black ink on ivory Japanese paper () which was priced, according to a note on the verso, at '21 guineas' (£22.1.0) and is now in the Hamburger Kunsthalle. Other dramatic impressions of the final state were printed in black ink on translucent Japanese paper (, ). However, Whistler worried about the condition of the copper plates at this late date:
20: Huish to Whistler, 8 June 1894, GUW #01270.
21: Whistler to Huish, 7 and 10 August 1894, GUW #01274, #01270, #13004.
'plates are as whimsical and as difficult to deal with as horses! And when I do take up these two or three remaining ones the Lord knows how they may jib and shy and buck and bolt with me - the trouble they will give I know well - and the amount of coaxing & light handling required before I get them back into anything like training again!'
In 1894 he still had 20 to print to complete the edition and in the next seven years only managed to print four impressions.
Eventually the Fine Art Society left him alone. Whistler's declining health made it increasingly unlikely that he would complete the editions. 23
23: Brown to Whistler, 15 November 1901, GUW #01370, and [16/23 November 1901, #01371.
The artist died in July 1903. Huish immediately wrote to Rosalind Birnie Philip
(1873-1958), asking for the Venetian copper plates. Eventually, it was arranged that Frederick Goulding
(1842-1909) should print the remaining 16 impressions of The Doorway
. Goulding completed the edition under her supervision by 14 December 1903. 24
Miss Philip recorded 'Dec. 3rd. "The Doorway". 24 proofs. 16 for Fine Art Society.' 25
This suggests that while giving the F.A.S. the 16 impressions they needed to complete the edition, she kept several herself, posssibly six, given that Goulding would have been perfectly entitled to keep a couple as well.
24: Huish to R. B. Philip, 1 Aug., 14 Sept., 22 and 29 Oct., 13 Nov., 3, 12, 14 and 17 Dec. 1903, GUL F323-4, 326; Philip to Huish, 15 Sept., 26 Oct., 18 Dec. 1903, GUL P485, F327-31, F333-35; MacDonald 2001[more], pp. 132-133.
25: Letter book Whistler LB6/261, Glasgow University Library.
Goulding signed his impressions such as the one reproduced above, 'F. Goulding. Printer / Dec. 4. 1903.' These last proofs, closely modelled on Whistler's final proofs, are fine examples of Goulding's work, dramatic in their use of ink tone to enhance the etched lines. They show that the plates had survived in fine condition, and draw to an end the epic history of Whistler's Venice sets.